Topic Ranking Evaluation on Bills (test)
0.36Kendall's τ (Fit)PROXANN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| PROXANNBackbone=Qwen-2.5-72B2025.07 | 0.36 | 0.31 | |
| PROXANNBackbone=Qwen-2.5-32B2025.07 | 0.34 | 0.3 | |
| PROXANNBackbone=Llama-3.1-70B2025.07 | 0.26 | 0.19 | |
| PROXANNBackbone=GPT-4o2025.07 | 0.21 | 0.29 | |
| PROXANNBackbone=Llama-3.1-8B2025.07 | 0.16 | 0.15 | |
| PROXANNBackbone=Qwen-2-7B2025.07 | 0.12 | 0.28 | |
| HumanProtocol=Leave-one-out inter-annotator correlation2025.07 | 0.09 | 0.18 | |
| NPMI2025.07 | 0.01 | -0.02 |